[Radiance-general] rsensor/rcontrib

[Greg Ward]

Hi Aris,

I am puzzled by your use of the -I+ option to rcontrib (first command) and -i option to rsensor (second command). These options are very different from each other, and I don't think you really want to use either.  The -I+ option tells rtrace to interpret its 6 input values as virtual illuminance point and orientation, rather than ray origin and direction as produced by rsensor.  Using the -i option in the second rsensor command tells the program to evaluate surfaces in your scene as if they were all Lambertian with 314% reflectance.  Again, probably not what you had in mind.

Try taking these options off and seeing what you get.  The rest of it looks OK as far as I can see.

Cheers,
-Greg

> From: Aris Tsangrassoulis <atsagras at arch.uth.gr>
> Subject: [Radiance-general] rsensor/rcontrib
> Date: April 29, 2014 9:43:44 AM PDT
> 
> Dear all,
> 
> trying to compare illuminance values on  a single point on the working surface located in the middle of a south facing room, three methods have been used (rtrace, rsensor and 3phase method using rsensor|rcontrib to estimate the view matrix) with the results difference been explainable. Sensor file represents spatial resonse file of a cosine corrected sensor.
> When I have changed the sensor file with one having a narrow filed of view (2x10 degrees) illuminance value using the 3phase approach was the same (~1% difference) as in the case of the cosine corrected sensor.
> The command was :
> rsensor -h -rd 5000  -vf view_test.vf  narrow10.dat . |rcontrib   -c 5000 -f klems_int.cal -bn Nkbins -b kbinS -m glazing -I+ -ab 11 -ad 50000 -ds .1 -lw 1e-5 test_r.oct>narrow10.vmx
> 
> Repeating the calculation with rsensor command only (again narrow versus cosine corrected sensor file):
> 
> rsensor -h -rd 5000 -ab 5 -ad 2500 -ar 300 -aa 0.1 -i -vf view_test.vf narrow10.dat  overcast.oct|rcalc -e "$1=179*($1*0.265+$2*0.670+$3*0.065)">overcast_rsensor_10.dat
> 
> results differ 98%. Any hint would be appreciated.
> I've examine the ray directions coming out from rsensor for both the sensor files and are in accordance with the spatial response. Using NREL's R4.2.a.4 on Windows.
> 
> Thanx,
> 
> Aris Tsangrassoulis